130 ROCHESTER ACADEMY OF SCIENCE. [April 1 8. 



If the object in view is to obtain a quantity of the heavier ele- 

 ment alone, a useful wrinkle is to combine this method of drawing 

 off with that of an alternating weaker and stronger current so as to 

 send into the tube b the grains of mean specific gravity along with 

 the lighter ones, the tube in- this case being lowered down to the level 

 at which, with the normal current, the lowest of the lighter grains 

 swim. In this way, for example, we have obtained nearly pure mona- 

 zite (sp. gr. 5) from a mixture with titaniferous iron (sp. gr. 4.75.) 



8. — With a fine adjustment of the current and a repetition of the 

 process on the parts already passed in the apparatus, the separation 

 may be made nearly perfect. As the withdrawal of material is per- 

 fectly under control and as uniformity of grain can be nearly obtained 

 by repetitions of the process, the rest depends on the skill of the 

 operator and particularly on the facility with which he recognizes the 

 different minerals in the dancing column of sand within the tube. 

 Naturally the composition of the sand operated upon and the relative 

 specific gravity of its different elements should be previously known 

 from a preliminary examination, and the movements of some mineral 

 that, from its well characterized color or aspect, can be readily 

 followed by the eye, should be taken as a guide. If the mixture does 

 not already contain such a characteristic guide mineral, it is advan- 

 tageous to add a few grains of one, which can afterwards be sep- 

 arated by the electro-magnet or by picking out under the lens. Gar- 

 net, ruttle, titaniferous iron, zircon and monazite give a good range 

 of color and specific gravity from which selection for this purpose 

 can be made, and, in any collection containing numerous samples of 

 sand, will almost always be at hand. 



To test the process, the heavy residue from a decomposed mus- 

 covite granite was passed once through the apparatus, withdrawn in 

 two portions, and the different minerals completely separated under 

 the lens and weighed. The sand was prepared by passing through a 

 Thoulet solution of about specific gravity 3, partially cleaned of iron 

 minerals with the electro-magnet and screened between bolting cloth 

 Nos. I and 4 (20 and 25 holes to a centimeter). The residue then con- 

 sisted principally of titaniferous iron, monazite and xenotime with a 

 few grains of staurolite, tourmaline and muscovite, the total quantity 

 being 1.77 grammes. The following table gives the percentages of 

 the three principal minerals in the original sample, in the lighter (I) 

 and heavier (II) portions, and those of the total amount of each 

 mineral in the two portions 



